25-Hydroxyvitamin D deficiency is associated with inflammation-linked vascular endothelial dysfunction in middle-aged and older adults

Abstract

We tested the hypothesis that vascular endothelial function, assessed by endothelium-dependent dilation, is related to serum vitamin D status among middle-aged and older adults without clinical disease, and that this is linked to inflammation. Brachial artery flow-mediated dilation (FMD), a measure of endothelium-dependent dilation, was lower (P<0.01) in vitamin D insufficient (3.7±0.2%; serum 25-hydroxyvitamin D [25(OH)D]: 20-29ng/mL, 62±1yr, n=31, mean±S.E.) and deficient (3.2±0.3%; 25(OH)D: <20ng/mL, 63±2 yr, n=22) vs. sufficient (4.6±0.4%; 25(OH)D: >29ng/mL; 61±1yr, n=22) subjects, whereas endothelium-independent dilation (brachial dilation to sublingual nitroglycerine) did not differ (P=0.45). Among all subjects, brachial FMD was positively related to serum 25(OH)D (%Δ: r=0.35, P<0.01), but not 1,25-dihydroxyvitamin D (r=-0.06, P=0.61), the active form of vitamin D. Vascular endothelial cell expression of the pro-inflammatory transcription factor nuclear factor κB (NFκB) was greater in deficient vs. sufficient subjects (0.59±0.07 vs. 0.44±0.05, P<0.05), and inhibition of NFκB (4d oral salsalate) improved FMD to a greater extent in subjects with lower vs. higher 25(OH)D (+3.7±0.6 vs. +2.0±0.2%, P<0.05). Endothelial cell expression of the downstream pro-inflammatory cytokine interleukin 6 also was higher in deficient vs. sufficient subjects (0.67±0.08 vs. 0.47±0.05, P<0.01) and inversely related to serum 25(OH)D level (r=-0.62; P<0.01), whereas vitamin D receptor and 1-α hydroxylase, the 25(OH)D to 1,25-dihydroxyvitamin D converting enzyme, were lower (P<0.05). Inadequate serum 25(OH)D is associated with vascular endothelial dysfunction among healthy middle-aged/older adults and this is mediated in part by NFκB-related inflammation. Reduced vitamin D receptor and 1-α hydroxylase may be molecular mechanisms linking vitamin D insufficiency to endothelial dysfunction.

Cardiovascular diseases (CVD) remain the leading cause of death in modern societies and advancing age is the major risk factor for CVD ^1-3. The increase in CVD risk with aging is attributable in large part to the development of vascular endothelial dysfunction, most commonly assessed as impaired endothelium-dependent dilation ^{2, 4, 5}. Consistent with this, middle-aged and older adults demonstrate reduced brachial artery flow-mediated dilation (FMD), a measure of endothelium-dependent dilation ⁶, compared with young adults ⁷. However, there is considerable variability in brachial artery FMD among middle-aged/older adults, and the contributing factors to impaired brachial FMD in this at risk group are incompletely understood.

Vitamin D deficiency is an independent predictor of CVD and all cause mortality ^8-10. Patients with chronic clinical diseases who are vitamin D deficient and young adults with severe vitamin D deficiency demonstrate impaired brachial artery FMD ^11-13. The physiological mechanisms by which vitamin D deficiency is linked to brachial artery FMD in these settings have not been established, but may involve vascular inflammation. In humans, vitamin D deficiency is associated with increased circulating inflammatory proteins ^{10, 14, 15}, whereas in cultured vascular endothelial cells vitamin D inhibits activation of the proinflammatory transcription factor, nuclear factor κB (NFκB) ¹⁶, as well as the release of the inflammatory cytokine, interleukin-6 (IL-6) ¹⁷, a downstream target of NFκB activation. However, it is unknown if vitamin D deficiency is associated with impaired brachial artery FMD among middle-aged/older adults without CVD and, if so, whether this is related to vascular inflammation.

In the present study, we tested the hypothesis that brachial artery FMD is inversely related to vitamin D status (as reflected by serum concentrations of 25-hydroxyvitamin D [25(OH)D]) among middle-aged/older adults in the absence of clinical diseases. We also tested the hypothesis that vitamin D deficiency and impaired brachial FMD are associated with vascular inflammation in this group. To do so, we determined brachial artery FMD in healthy middle-aged/older men and women differing in serum 25(OH)D levels. We then assessed the expression of NFκB and IL-6 in vascular endothelial cells obtained from subgroups of these subjects, as well as NFκB signaling-associated suppression of brachial artery FMD using salsalate, an inhibitor of NFκB activation. To provide additional insight into the molecular mechanisms linking circulating 25(OH)D to endothelial function, we measured 1-α hydroxylase, the enzyme that converts 25(OH)D to 1,25-dihydroxyvitamin D (1,25(OH)D₂ (the biologically active form of vitamin D) and vitamin D receptor in endothelial cells obtained from subgroups of vitamin D sufficient and deficient subjects.

Methods

Subjects

Subjects were 75 middle-aged/older (50-79 years) men (n=47) and postmenopausal women (n=28) (92% Caucasian, 5% Asian, 3% Hispanic). Subjects had a systolic blood pressure <159 mmHg, diastolic blood pressure <99 mmHg and were otherwise free of CVD, diabetes, kidney disease and other chronic clinical disorders as assessed by medical history, physical examination, blood chemistries and resting and exercise ECG. All subjects were non-smokers, not taking medications (prescription or over the counter), hormone replacement therapy or dietary supplements (including those with antioxidant properties). Subjects were not performing regular aerobic exercise (i.e. <30 min/day, <2 days/week for at least 2 years prior to study participation). All procedures were approved by the Human Research Committee of the University of Colorado at Boulder. The nature, benefits and risks of the study were explained to the volunteers and their written informed consent was obtained prior to participation.

Study Procedures

All measurements were performed at the University of Colorado at Boulder Clinical and Translational Research Center (CTRC) after an overnight fast (water only) and a 24-hour abstention from alcohol and vigorous physical activity.

Subject Characteristics, Blood Assays and Dietary Analysis

Subjects were classified by 25(OH)D concentration as deficient (<20 ng/mL), insufficient (20-29 ng/mL) or sufficient (>29 ng/mL) in accordance with recent recommended guidelines ¹⁸. See http://hyper.ahajournals.org for details of subject characteristics, blood analyses and dietary analysis.

Brachial Artery FMD and Endothelium-Independent Dilation

Brachial artery FMD (upper forearm cuff position), peak shear rate during FMD and endothelium-independent dilation (brachial artery dilation in response to sublingual nitroglycerin) were determined using duplex ultrasonography (Power Vision 6000, Toshiba) with a linear array transducer as described previously ^{19, 20}. Responses are expressed as mm and percent change from baseline diameter per recent recommended guidelines ²¹.

Vascular Endothelial Cell Protein Expression

Vascular endothelial cells were obtained from antecubital venous sampling and analyzed for expression of proteins (IL-6, NFκB p65, TNFα, vitamin D receptor or 1-α hydroxylase) using quantitative immunofluorescence as described previously ^{20, 22-24}. Inter-individual and group differences in protein expression using this procedure reflect expression in endothelial cells obtained from brachial artery sampling in the same subject ²⁵. See http://hyper.ahajournals.org for details of the methodology.

Salsalate Administration

To determine the contribution of NFκB-associated pro-inflammatory signaling to 25(OH)D-related differences in EDD, a sub-group of 21 middle-aged/older (50-68 years) men (n=17) and postmenopausal women (n=4) were randomly assigned to receive oral doses of salsalate, a NFκB inhibitor, or placebo using a 4-day cross-over design as described previously ²⁰. See http://hyper.ahajournals.org for details.

Statistics

All of the data are presented as means±S.E.M. See http://hyper.ahajournals.org for details on statistical analysis.

Results

Subject Characteristics

General subject characteristics, blood pressure, metabolic factors and kidney function are presented in Table 1. Mean serum 25(OH)D in the overall sample was 25.3±1.0 ng/mL (range 7-47): 29% of subjects were 25(OH)D deficient and this was similar in men vs. women (30% and 29%). The 25(OH)D deficient, insufficient and sufficient groups did not differ significantly in age, percent body fat, hip circumference, waist-to-hip ratio, physical activity, total bone mineral density, systolic and diastolic blood pressure, plasma lipids and lipoproteins, plasma glucose and insulin, HOMA, eGFR or 1,25(OH)₂D, whereas there were selective group differences in body mass, BMI and waist circumference (P<0.05). The season in which blood samples were collected (not shown) was only weakly related to serum 25(OH)D (eta coefficient: 0.29) and 1,25(OH)₂D (eta coefficient: 0.16) in the overall sample.

Table 1.

Subject Characteristics	Deficient (<20 ng/mL)	Insufficient (20-29 ng/mL)	Sufficient (>29 ng/mL)
n (m/f)	22(14/8)	31 (21/10)	22 (12/10)
25(OH)D (ng.mL-1)	15.8±1.2	24.6±0.5*	35.6±1.2 †‡
1,25(OH)2D (pg.mL-1)	46.5±3.2	50.0±2.4	48.2±4.6
Age (years)	63±2	62±1	61±1
Body Mass (kg)	83±4	82±3	73±2 §
BMI (kg/m2)	28.0±1.0	27.1±0.75	25.1±0.75 *
Total Body Fat (%)	35±2	33±1	30±2
Waist Circumference (cm)	95±3	92±3	85±3 *
Hip Circumference (cm)	105±2	104±2	100±2
Waist:Hip Ratio	0.91±0.02	0.89±0.02	0.85±0.02
PA (MET hrs.wk-1)	17±4	24±4	22±4
Total BMD (g/cm-1)	1.20±0.03	1.24±0.02	1.25±0.03
Systolic BP (mmHg)	129±3	127±3	122±3
Diastolic BP (mmHg)	78±2	76±1	74±2
Total Cholesterol (mg.dL-1)	209±7	204±5	198±7
LDL Cholesterol (mg.dL-1)	127±7	127±4	121±5
HDL Cholesterol (mg.dL-1)	57±4	55±2	57±3
Triglycerides (mg.dL-1)	127±13	111±10	104±10
Plasma Glucose (mg.dL-1)	95±3	93±2	91±2
Insulin (μU.L-1)	7.6±1.6	7.0±0.90	6.1±0.68
HOMA	2.00±0.49	1.62±0.23	1.43±0.18
eGFR (mL.min-1.(1.73 m2) -1)	83±3	81±3	89±3

Data are mean ± S.E.

^*p<0.05 vs. deficient;

^†p<0.01 vs. deficient;

^‡p<0.01 vs. insufficient;

^§p=0.05 vs. insufficient.

25(OH)D, 25-hydroxyvitamin D3; 1,25(OH)₂D, 1,25-dihydroxyvitamin D3; BMI, body mass index; PA, physical activity (average daily leisure and occupational activity); MET, metabolic equivalent; BMD, bone mineral density; BP, blood pressure; LDL, low-density lipoprotein; HDL, high-density lipoprotein; HOMA, insulin sensitivity index; (homeostasis model assessment); eGFR (estimated glomerular filtration rate, Modification of Diet in Renal Disease (MDRD) Study Equation).

Humoral Factors

Humoral factors are shown in Table 2. C-reactive protein, TNF-α, oxidized LDL, total antioxidant status, endothelin-1 and norepinephrine did not differ among the groups. Plasma IL-6 was higher in the deficient group (p<0.01 vs. sufficient) and was inversely related to 25(OH)D the overall sample (r = -0.40) (P<0.01).

Table 2.

Humoral Factors	Deficient (<20 ng/mL)	Insufficient (20-29 ng/mL)	Sufficient (>29 ng/mL)
CRP (mg.L-1)	1.3±0.3	1.7±0.3	1.0±0.3
IL-6 (pg.mL-1)	1.8±0.2	1.4±0.2	1.1±0.1 *
TNF-α (pg.mL-1)	1.5±0.1	1.3±0.1	1.7±0.3
Oxidized LDL (U.L-1)	53.3±3.6	62.6±2.6	58.1±2.7
Total Antioxidant Status (mmol.L-1)	1.19±0.04	1.22±0.03	1.24±0.06
ET-1 (pg.mL-1)	6.5±0.5	6.3±0.2	7.0±0.9
Norepinephrine (mg.dL-1)	311±28	311±31	275±28

Data are mean ± S.E.

^*p<0.01 vs. deficient.

CRP, C-reactive protein; IL-6, interleukin-6; TNF-alpha, tumor necrosis factor-α; LDL, low-density lipoprotein; ET-1, endothelin-1.

Diet Composition

Dietary analysis is shown in Table S1. Total energy intake, % macronutrient intake, and cholesterol, vitamin C, calcium, potassium and sodium intake were similar in the 3 groups. Mean values for dietary vitamin D tended to be progressively greater in the 25(OH)D deficient, insufficient and sufficient groups (P=0.20), and tended to be related to serum 25(OH)D in the overall sample (r=0.31, p=0.06).

Brachial Artery FMD and Endothelium-Independent Dilation

Baseline brachial artery diameter and peak shear rate during FMD did not differ among groups (see http://hyper.ahajournals.org) or relate to serum 25(OH)D level (P≥0.30). Brachial artery FMD was lower in the 25(OH)D deficient and insufficient vs. sufficient groups (-30% [%Δ] and -29% [mmΔ] deficient vs. sufficient) (P<0.05, Figure 1). Endothelium-independent dilation did not differ among groups (P≥0.40) (Please see http://hyper.ahajournals.org). In all subjects, brachial artery FMD was positively related to serum 25(OH)D level (%Δ: r=0.35, P<0.01; mmΔ: r=0.29, P=0.01) (Figure 1). This relation remained significant after correcting for subject characteristics in Table 1 (%Δ: R²=0.18, P<0.01; β=0.43, SE=0.03, P<0.01; mmΔ: R²=0.25, P<0.01; β=0.40, SE=0.002, P<0.01). Season of measurement only weakly related to FMD (eta coefficient: 0.19) and the relation between FMD and serum 25(OH)D was unaffected after adjusting for season of measurement (partial correlation coefficient: (%Δ r=0.36, P<0.01; mmΔ r=0.30, P<0.01).

Figure 1.

Brachial FMD (percent change (%Δ), top; absolute change (mmΔ); bottom) in 25(OH)D deficient, insufficient and sufficient groups (left); relation between serum 25(OH)D levels and FMD in the overall sample (%Δ, top; mmΔ, bottom) (right). Values are mean ± S.E. *P<0.01 vs. deficient; †P<0.05 vs. insufficient; ‡P<0.05 vs. deficient.

In contrast to 25(OH)D, brachial artery FMD did not differ across tertiles of serum 1,25(OH)₂D and was not related to serum 1,25(OH)₂D in the overall sample (%Δ r= -0.06, P=0.61; mmΔ r= -0.06, P=0.61) (see http://hyper.ahajournals.org). Endotheliumin-dependent dilation also did not differ across tertiles of serum 1,25(OH)₂D levels (P≥0.60).

Vascular Endothelial Cell NFκB and Brachial Artery FMD

Expression of NFκB was greater in vascular endothelial cells of the 25(OH)D deficient compared with the sufficient group (P<0.05, left Figure 2).

Figure 2.

Protein expression of NFκB p65 in vascular endothelial cells from 25(OH)D deficient, insufficient and sufficient subjects (left). FMD (%Δ) (center) and protein expression of NFκB p65 (right) in vascular endothelial cells following placebo vs. salsalate in lower and higher 25(OH)D groups .

When subjects were divided by the median value of 25(OH)D, brachial FMD was 42% lower in the group with lower compared with higher circulating vitamin D (P<0.05, middle Figure 2). NFκB inhibition with salsalate improved FMD in both groups (P<0.05 vs. placebo), but the increase was greater (P<0.05) in subjects with lower (+168%) compared with higher (+50%) serum 25(OH)D. As a result, FMD was similar in the 2 groups following treatment with salsalate. There were no differences in baseline diameter or shear rate between groups or conditions (salsalate vs. placebo). Groups did not differ in any subject characteristics, and serum 25(OH)D concentrations did not differ between placebo and salsalate conditions. The results were similar when mmΔ was used instead of %Δ.

Salsalate treatment reduced the expression of NFκB in vascular endothelial cells obtained from subjects with lower (P=0.06), but not from those with higher serum 25(OH)D (right Figure 2). There were no differences in endothelial cell expression of NFκB in the 2 groups following treatment with salsalate.

Vascular Endothelial Cell IL-6 and TNFα Expression

25(OH)D-deficient subjects had 81% greater endothelial cell protein expression of IL-6 compared with sufficient subjects (P<0.01, Figure 3). IL-6 was inversely related to 25(OH)D among all subjects (r=-0.62, P<0.01). In contrast, endothelial cell expression of TNFα did not differ among groups (P=0.60) (Figure 3) or correlate with 25(OH)D (r=-0.03, P=0.86).

Figure 3.

Protein expression of IL-6 (left) and TNFα (right). Relation between 25(OH)D and IL-6 (middle). For all proteins, values are venous endothelial cell expression relative to human umbilical vein endothelial cell (HUVEC) control. *P<0.05 vs. deficient/lower; †P<0.01 vs. deficient/lower; P=0.06 vs. lower 25(OH)D group placebo condition.

Vascular Endothelial Cell Vitamin D Receptor and 1-α Hydroxylase Expression

Serum 25(OH)D-deficient subjects had lower endothelial cell protein expression of vitamin D receptor compared with sufficient subjects (P<0.05, Figure 4). Endothelial cell protein expression of 1-α hydroxylase also was lower in serum 25(OH)D deficient vs. sufficient subjects (P<0.05) and was positively related to FMD (%Δ r= 0.82, P<0.01; mmΔ r= 0.92, P<0.01) (Figure 4).

Figure 4.

Protein expression of vitamin D receptor (VDR) (left) and 1-α hydroxylase (1-OHase) (middle) in vascular endothelial cells obtained from 25(OH)D deficient and sufficient subjects; representative images shown below. Relation between endothelial cell protein expression of 1-OHase and FMD (mmΔ) (right). For all proteins, values are venous endothelial cell expression relative to human umbilical vein endothelial cell (HUVEC) control. *P<0.05 vs. deficient.

Discussion

The findings of the present study are the first to show that vascular endothelial function is related to vitamin D status (as reflected by serum 25(OH)D) among middle-aged/older adults free of clinical disease. Our results also provide the first evidence in humans that lower serum 25(OH)D is associated with vascular endothelial inflammation, increased NFκB signaling-related suppression of vascular endothelial function, and reduced vascular endothelial vitamin D receptor and 1-α hydroxylase expression. In contrast to 25(OH)D, our data indicate that brachial artery FMD is not related to serum 1,25(OH)₂D in this group.

Vitamin D and Vascular Endothelial Dysfunction with Aging

The results of the present study demonstrate that brachial artery FMD, a measure of endothelium-dependent dilation and vascular endothelial function, is inversely related to serum vitamin D status among middle-aged/older adults without clinical disease. This is consistent with recent work in patients with diabetes and chronic kidney disease ^{11, 12} and in young adults with severe vitamin D deficiency ¹³. Our findings extend these observations by showing an inverse association between circulating vitamin D and vascular endothelial function that is evident in healthy adults within a typical range of serum 25(OH)D ²⁶.

Our data indicate a specific relation between brachial artery FMD and serum 25(OH)D, the metabolite used to determine vitamin D status, but not 1,25(OH)₂D, the biologically active form of vitamin D. 1,25(OH)₂D is formed via conversion of 25(OH)D by the enzyme 1-α hydroxylase, a process historically viewed as occurring in the kidney with the presumed aim of maintaining muscle and bone health. However, recently it has been recognized that this conversion also can occur locally in a number of different cells and tissues ^{18, 27}. This implies that 1,25(OH)₂D can exert autocrine/paracrine cell-specific functions when produced outside of the kidney ²⁸.

In the context of the present work, vascular endothelial cells in culture express vitamin D receptors and 1-α hydroxylase ^{29, 30}. Here we demonstrate for the first time expression of these proteins in vascular endothelial cells collected from human subjects. Because circulating 1,25(OH)₂D did not differ between groups and was not related to brachial FMD, our results suggest that the association between serum 25(OH)D level and FMD could be mediated in part by vascular endothelial cell conversion of 25(OH)D to 1,25(OH)₂D. That endothelial cells from vitamin D deficient subjects had reduced expression of vitamin D receptors and 1-α hydroxylase compared with vitamin D sufficient subjects and that 1-α hydroxylase was strongly related to FMD are consistent with this possibility. Thus, reduced expression of vitamin D receptors and 1-α hydroxylase in vitamin D deficient subjects may have limited the conversion of 25(OH)D to 1,25(OH)₂D and attenuated vitamin D signaling in the vascular endothelium, thus contributing to reductions in FMD. If so, this represents a novel mechanism by which circulating vitamin D deficiency may be linked to vascular endothelial dysfunction.

It should be noted that earlier investigations in patients with chronic kidney disease on hemodialysis and healthy young adults with severe vitamin D deficiency found that brachial FMD was related to both serum 25(OH)D and 1,25(OH)₂D levels ^{11, 13}. The differences between our results and those of previous investigations on this point may be explained by differing kidney function (i.e. kidney disease vs. healthy) and/or the degree of vitamin D deficiency. Differences in oxidative stress between vitamin D sufficient and deficient groups also may have been greater in these previous studies, as reflected by increased circulating markers of oxidant damage ¹³. In contrast, plasma oxidized LDL, a marker of oxidant modification of lipids, did not differ among the groups in the present study.

Vitamin D and Vascular Inflammation

Our results are the first to show a relation between serum vitamin D status and vascular endothelial inflammation in humans. Specifically, we found that vascular endothelial cell expression of the p65 subunit of NFκB, a major pro-inflammatory nuclear transcription factor, and IL-6, a pro-inflammatory cytokine and downstream target of NFκB, were greater in vitamin D deficient compared with sufficient middle-aged/older adults. Indeed, IL-6 expression in endothelial cells was strongly inversely related to serum 25(OH)D. In contrast, the expression of TNFα, a pro-inflammatory cytokine and (upstream) activator of NFκB, was not associated with 25(OH), suggesting no obvious involvement in vitamin D-related effects on endothelial cell inflammation. Previous investigations have reported a positive correlation between circulating pro-inflammatory cytokines and serum 25(OH)D levels ^{10, 14, 15}, and that vitamin D inhibits experimental activation of NFκB, as well as IL-6 release, in endothelial cell culture ^{16, 17}. Our findings extend this prior work to show an inverse association between serum 25(OH)D level and pro-inflammatory NFκB-IL-6 activation in vascular endothelial cells obtained from human subjects.

To determine if differences in NFκB-related signaling in the vascular endothelium between subjects with lower compared with higher 25(OH)D may have contributed to differences in vascular endothelial function, we inhibited NFκB activation using short-term treatment with salsalate ²⁰. Salsalate reduces endothelial cell NFκB in adults with elevated baseline expression ²⁰. Consistent with this, in the present study we observed a selective reduction in NFκB in endothelial cells obtained in subjects with lower serum 25(OH)D concentrations, a group with elevated baseline expression of NFκB. Most importantly, we found that the tonic suppression of EDD by NFκB signaling, as assessed by the improvement in brachial artery FMD from placebo control in response to short-term treatment with salsalate, was greater in the subgroup with lower circulating 25(OH)D, and that salsalate treatment abolished group differences in FMD. This provides direct evidence of NFκB activation-related vascular endothelial dysfunction in middle-aged/older adults with lower serum 25(OH)D.

Limitations

In the present study, selective subject characteristics were or tended to be different between groups. However, these trends did not explain group differences in brachial FMD, as correcting for these factors did not weaken the relation between FMD and serum 25(OH)D concentrations. Parenthetically, the tendency for differences in such characteristics in the present study among healthy adults is consistent with evidence in human studies and animal models (e.g., vitamin D receptor knockout mice) that vitamin D deficiency is associated with cardiovascular risk factors such as increased blood pressure and insulin resistance ^{31, 32}.

Expression of proteins, including NFκB, in endothelial cells obtained from the antecubital vein generally reflects expression in cells obtained from the brachial artery of the same subject ^{22, 25}. However, in the present study we assessed only cells from venous sampling and, thus, cannot confirm the presence of differences in arterial endothelial cells among groups differing in serum vitamin D status.

Last, we did not measure for proteinuria, and kidney function was estimated by the MDRD equation, which is less accurate in individuals with normal or mildly impaired kidney function compared with patients with more severe dysfunction. Therefore, we cannot rule out differences in kidney function among the three serum 25(OH)D groups.

Conclusions

In conclusion, the results of the present study support the hypothesis that serum 25(OH)D status is associated with vascular endothelial function among middle-aged/older adults in the absence of clinical disease. Our findings also demonstrate that lower 25(OH)D status is associated with increased vascular endothelial cell expression of NFκB and IL-6, and increased NFκB-related suppression of vascular endothelial function. Our data indicate that, in contrast to 25(OH)D, FMD is not associated with serum 1,25(OH)₂D in this group. Finally, our results show for the first time that vascular endothelial cell expression of vitamin D receptor and 1-α hydroxylase are decreased with 25(OH)D deficiency and are related to vascular endothelial function.

Perspectives

These observations provide an experimental basis for intervention trials aimed at determining the efficacy of vitamin D treatment for reversing vascular endothelial dysfunction in 25(OH)D deficient and insufficient middle-aged/older adults to prevent cardiovascular diseases, and the role of inhibition of inflammatory signaling in any such benefits.